RAN Resource Sharing for Low-Latency AI/ML Computation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for resource allocation in radio access networks (RAN) are inefficient, particularly for computationally intensive tasks like AI/ML, due to limited computational power in user devices and slower processing times, leading to reduced efficiency and longer convergence times.
Innovation Solution
Enhanced protocols for RAN nodes to share computational resources horizontally by reporting capabilities and requesting assistance from neighboring base stations, eliminating the need for a central controller and reducing data transmission latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If computational tasks are performed locally at user devices with limited computational power, then device autonomy is maintained, but processing time increases and efficiency decreases
Solution Approach 1:
The patent combines computational tasks across multiple user devices and base stations into a distributed computing pool. By merging the computational resources of multiple nodes (user devices and base stations) into a unified resource allocation system, the system achieves higher overall computational efficiency while reducing individual processing times through collaborative computation.
Solution Approach 2:
The patent introduces a new dimension to resource allocation by considering both local device capabilities and network-wide resource availability. The enhanced protocols enable base stations to request computational assistance from neighboring base stations, creating a multi-dimensional resource sharing architecture that goes beyond traditional single-node processing.
2Ease of operation
If a central controller manages resource allocation, then coordination is simplified, but data transmission latency increases
Solution Approach 1:
The patent extracts the resource allocation function from a centralized controller and distributes it to individual base stations. Each base station autonomously manages its computational resources and coordinates with neighboring base stations through enhanced protocols, eliminating the bottleneck of centralized control while maintaining coordination through distributed decision-making.
Solution Approach 2:
Base stations are enabled to self-manage their computational resources by reporting capabilities and requesting assistance from neighboring base stations independently. This self-service mechanism allows each base station to autonomously optimize resource allocation without requiring constant central controller intervention, thereby reducing transmission latency.
3Device complexity
If computational resources are allocated vertically from application layer to RAN, then hierarchical control is maintained, but resource utilization efficiency decreases
Solution Approach 1:
The patent segments the computational resource allocation process into independent horizontal interactions between neighboring base stations rather than a single vertical hierarchical flow. Each base station independently reports its computational capabilities and requests resources from peers, creating multiple parallel allocation paths that improve overall resource utilization efficiency.
Solution Approach 2:
The patent introduces dynamic resource allocation where base stations can request computational assistance from neighboring base stations based on real-time needs and available resources. This dynamic horizontal coordination allows the system to adapt resource distribution flexibly without rigid hierarchical constraints, improving allocation efficiency.
Data Source
AI summary
Wireless communication methods enhance computations in a radio access network (RAN). An example method involves a first RAN node receiving an interface setup message from a second RAN node, which includes resource capacity information. The interface setup message may relate to setting up a node-to-node interface between the first and second RAN nodes. The first RAN node transmits a resource status request and receives a resource status response containing resource availability information. This method allows for efficient resource management by utilizing neighboring cell site resources, thereby reducing data transmission capacity and latency.